help me not to use eval in a nested loop?

Hi I have a requiremen to import 15 Excel files which I managed, but still cant save the output

files share the same name almost, as follows: Speed(X)_Mvr(Y), herer X is [0.214, 0.252, 0.500 ], and Y is 1 to 5, so the names are like: Speed0.214_Mvr1, Speed214_Mvr2,......Speed0.5_Mvr5 = total is 15 files.

To read the files I made a nested loop that uses "sprintf" function to read every thing using xlsread: code is:

S1= [0.214, 0.252, 0.5];  %these are my speeds

for i=1:3                 %this loops my speeds
    for j=1:5             %this loops my Mvr values

       data = xlsread( sprintf('Speed%d_Mvr%d.csv',S1(i),j));

      end    
  end

I prefer to have the data saved as 15 diffrent Matrices in my Workspace to use later with somesort of identification reagrding Speed value and Mvr value. this can be done with eval, but im not allowed to for this assignment. all ideas welcome

Thanks

Answer 1

[Star Strider]

I assume that ‘can’t save the output’ means that you’re overwriting your ‘data’ variable each time. I would use a cell array:

data{i,j} = xlsread( sprintf('Speed%d_Mvr%d.csv',S1(i),j));

[Marwan Al]

Thanks Star Strider, Yes you are correct I am over writing the 'data' variable. And this Cell method works if I want to refer to each file with i, j location...But would like to name each elemnt in the cell if possible, like Speed1_Mvr1..etc.and be able to refer to such name during other calculations. Reason is each file is 3000×11 Matrix and I need to extract specific points and columns. In short Matlab reads 15 matrix files and outputs 15 matrices each with different name. Thanks again

[Star Strider]

My pleasure.

I don’t know precisely what you’re doing, but if all your matrices are the same size, you can use the cell2mat function to convert each cell to a double matrix, and then use the cat function to concatenate them into a 3D array.

Specifically, in the loop, you can use M = cat(cell2mat(X),3); where ‘X’ is the 2D matrix read from your Excel files. You can then address every element of the resulting double matrix ‘M’ as you wish. (Name them whatever you wish.) You would have to keep track of which ‘page’ of the 3D matrix corresponds to a specific file, so I leave that to you to organise.

I apologise for the delay in responding. Today is ‘errand day’, and I was away for two hours. (Life intrudes!)

[Walter Roberson]

http://matlab.wikia.com/wiki/FAQ#How_can_I_create_variables_A1.2C_A2.2C....2CA10_in_a_loop.3F

Answer 2

[John BG]

Hi, regarding naming each matrix, try

B={};for i=1:1:10 B{i}=['Speed' char(i+48) '_Mvr' char(i+48)]; end

then attach each B(1) B(2) .. to each Excel extracted matrix

[Marwan Al]

Thanks John I'll give that a try tomorrow and report back..

[Marwan Al]

Ok guys, kind of still stuck with this, @ John I didnt know how to attach names for elemets inside a cell, but to clarify my question, this is my current code using eval:

          S= [204 242 4];

          for i=1:3
              for j=1:5
               data = xlsread( sprintf('Dataset_0.%d_u%d.csv',S(i),j));
               eval(['Speed' num2str(S1(i)) 'Mvr' num2str(j) ' = data;'])    
              end    
          end

the output is 15 entries in the Workspace likle in picture

How can I achive the(( very same outputs )) without using eval?? Thank you all

[Stephen23]

@Marwan Al: to achieve that you can't avoid eval because eval is the only way to dynamically assign variable names like that. However dynamically named variables are a really slow, buggy and awful way to program (that beginners think is wonderful), so whoever told you to avoid eval is doing you a favor.

The best advice is: do not use eval. Learn to program using faster and more reliable methods, exactly like in John's answer.

Read this to know why using eval is such a poor way to program:

http://www.mathworks.com/matlabcentral/answers/196952-how-to-create-new-variables-in-batches-with-strcat